NavX broken (physically). What do we do?

Our navx was damaged. This is how it looks now:

What do we do? Is it broken beyond repair or can we try to resolder it?

1 Like

You can definitely try to resolder it. There’s a reasonable chance it’ll still work.

2 Likes

Have to say it could be worse

You’ll need to fix the pad that got torn off and resolder the rest. Not a job for an amateur, but certainly not beyond repair

9 Likes

The SPI enable trace is damaged, so there is no SPI.

The UART pads could be soldered together with a jumper wire. This would enable UART communication.

Working communication protocols would be USB, UART, and i2c.

1 Like

Time to learn SMD reflow.
It can be repaired, but it will be delicate work.

If repairing it yourself seems like too daunting of a task (or you don’t have the tools readily available), you could try reaching out to a local electronics/mobile phone repair shop. I’ve had a local shop do some similar personal work for me (non FRC related) on some prototype PCB’s before I got some of my own equipment.

Heck, mention that you’re with a robotics team, and maybe they’ll think that’s super cool and discount/donate the repair :woman_shrugging:

2 Likes

The part that broke off is called a “DIP switch” (Dual Inline Pin – a physical form factor or package). I don’t see any other damage, but it’s hard to tell for certain. But, it’s worth proceeding as though this is the only problem and then testing it.

Here’s the section of the manual that talks about these switch settings:

The part that’s missing either opens the pairs of pins, or shorts them together. With the part missing, both sets of pins are open. You can easily solder a small wire to (more-or-less) permanently short each pair of pins. Where there is a missing pad, you can try to trace back to where that pad is/was connected and solder there if you need to. If you can expose a little bit of copper where there is a line that was connected to the trace, you might be able to find the other end with an ohm-/multi- meter, or you might be able to follow it visually.

I don’t have access to a navX right now to try to track it down, but another option is asking someone who has a good one to trace this connection for you, using a meter.

When the sliders were in the “ON” position, the two pads in that row were shorted together – the wire mentioned above. When in the off position, they were not shorted – what you have now, with no wire.

It looks as though the default settings here are for both switches to be on. The damaged pad is on the SPI switch, which is the one that’s likely to be a little harder to fix – and the one you might
most want to fix.

If you don’t do anything to fix it, you should still be able to set things up in your code to use I2C. In other words, there’s a good chance things will either just work or that you can make a small code change to get things to work.

1 Like

A reasonably competent solderer with some experience doing SMD work could do that. One of the pads has been torn off (the top right).

If this was on my bench right now, I’d solder the switch back onto the remaining three pads after cleaning the contacts on the switch and applying a bit of flux. Then I’d scrape the solder mask off the trace previously connected to that pad and plop down a short strand of 32ga wire from the trace to the contact on the switch that would normally attach to that pad. Then I’d dribble some UV cure epoxy on the whole kitten-kaboodle and cure it with my UV pen-light. (if you don’t have 32ga or similarly thin wire then a strand of shielding from some coaxial cable will also work in a pinch)

Or, you might decide to follow @JamesT’s advice if you don’t care about SPI.

6 Likes

SPI is, in my experience, the only reliable way to communicate to the NavX. I strongly suggest a repair attempt given the relatively minor extent of the damage.

3 Likes